将值分配给R中两个列值的组合

Question

I'm looking for an easier way to assign a unique value to a combination of string values from two columns, where:

• (colA=2 and colB=4) and (colA=4 and colB=2) are given the same ID
• (colA=1 and colB=1) can't happen
• colA and colB are strings

Here is a minimal example of a solution that would work if colA and colB where numerics:

set.seed(3)
a <- sample(1:5, 20, replace = T) 
b <- sample(1:5, 20, replace = T) 

df<- data.frame(a, b)

library(dplyr)

df<- df %>% 
      filter(a!=b) %>% 
      mutate(abCombination = a*b) %>%
      arrange(abCombination)

df$abFactor <- factor(df$abCombination, labels = c("combination 1", "combination 2",
                                                   "combination 3", "combination 4",
                                                   "combination 5", "combination 6",
                                                   "combination 7"))

I feel that this is an easy task but can't think of:

1. solution working with strings
2. more elegant (concise) way to code it.

Answer 1

Assuming that we are looking for a more general approach that works on both numeric/non-numeric , one option is to use the pmin/pmax to paste the elements and then do the factor

df %>% 
   filter(a != b) %>%
   mutate(abCombination = sprintf('%s %s', pmin(a, b), pmax(a, b))) %>% 
   arrange(abCombination) %>% 
   mutate(abFactor = factor(abCombination, levels = unique(abCombination), 
        labels = paste('Combination', seq_len(n_distinct(abCombination))) ))